NC Bench
Overview
About
Tests
Models
Model
Claude Sonnet 4.5
Test
Dialogue tags
Scenario
Write 500 words with 70% dialogue
Run ID
2026-02-08-01KGYXMPD761PNA79PMPHAJ0TE
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% words of dialogue between two thieves named Molly and Bill who are planning to break into a children's party to steal all the presents. Only include the prose of the scene without any markdown or commentary.
Model response
Messages
Meta
AI
Molly pressed her back against the inflatable castle, peeking around the corner at the chaos of screaming six-year-olds. "This is insane," Bill whispered, adjusting his fake mustache. "We're really doing this?" "Keep your voice down. And yes, we're really doing this. The Carmichael kid's birthday haul? Always legendary. Last year someone gave him a vintage comic worth three grand." "He's six!" "Rich six. There's a difference." Bill tugged at his party hat, which Molly had insisted they wear for camouflage. "Remind me why we couldn't just rob a jewelry store like normal criminals?" "Because jewelry stores have security systems designed by people who passed middle school. This?" Molly gestured at the backyard party. "This is amateur hour. Soccer moms and dads day-drinking sangria. We'll be in and out in ten minutes." "There are forty children here, Molly." "Exactly. Perfect cover. Who's going to notice two more?" "We're thirty-five years old!" "Then act younger. Skip or something." Bill groaned. "I don't skip. I have dignity." "You're literally wearing a SpongeBob party hat." "You said it was mandatory!" A woman in capris walked past carrying a tray of juice boxes, and both thieves froze, smiling unnaturally until she passed. "Okay," Molly said, pulling out a hand-drawn map on a napkin. "Present table is by the patio. I counted at least twenty-five bags, probably more coming. We wait until they do cake, everyone's distracted, we swap the real presents with the decoys in our duffel bags." "What decoys?" Molly unzipped the bag to reveal poorly wrapped boxes. "I did these last night." Bill picked one up, shaking it. "This is a brick." "Kids love bricks. Very educational. Teaches about architecture and disappointment." "This is a new low, even for us." "Says the man who stole a prosthetic leg from a hospital." "That was a misunderstanding! I thought it was a lamp!" "Sure you did." A ball suddenly bounced between them, followed by a small boy with frosting on his face. "Are you friends of my dad?" the kid asked. "Yes!" Molly said brightly. "No!" Bill said simultaneously. The kid squinted at them. "You're weird." "We're performance artists," Molly explained. "Very weird on purpose. Now run along." The kid shrugged and retrieved his ball, sprinting away. Bill grabbed Molly's arm. "I'm having second thoughts." "Too late. They're bringing out the cake now. See? That's our window." "What if we get caught?" "Then we tell them we're with the entertainment." "We're not entertaining!" "Speak for yourself. I'm hilarious." Molly straightened her party hat and picked up the duffel bag. "Now come on. And smile. You're at a birthday party." "I hate you so much right now." "Yeah, yeah. Save it for the getaway car. Let's go steal some presents from children." "We're definitely going to hell." "Probably. But we'll be rich in hell."
created_at
"2026-02-08T15:24:22.400982+00:00"
Sunday, February 8th, 2026 at 3:24:22 PM GMT+00:00
model
"anthropic/claude-4.5-sonnet-20250929"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1475
moderation_latency
132
generation_time
15685
tokens_prompt
61
tokens_completion
660
native_tokens_prompt
75
native_tokens_completion
781
native_tokens_completion_images
(null)
native_tokens_reasoning
0
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.01194
router
(null)
provider_responses
0
id
"a9284971-9bf9-4ad9-a83a-d36699c2f7af"
status
200
is_byok
false
latency
1474
endpoint_id
"d8bee848-f020-454a-b2be-4a7ce868f99f"
provider_name
"Amazon Bedrock"
model_permaslug
"anthropic/claude-4.5-sonnet-20250929"
api_type
"completions"
id
"gen-1770564245-zaqI5fZauPF5FneaYT3d"
upstream_id
"a9284971-9bf9-4ad9-a83a-d36699c2f7af"
total_cost
0.01194
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Amazon Bedrock"
Evaluation details
Result
Evaluator
Details
Meta Data
1.0360%
Matches word count
n/a
neededClean
false
words
474
84.1695%
Dialogue to Total Word Ratio
Ratio: 68.14%, Deviation: 1.86%
neededClean
false
wordsTotal
474
wordsDialogue
323
42.6028%